Paracetamol Accelerates Closure of the Ductus Arteriosus after

Premature Birth: A Randomized Trial

Pia H€arkin, MD, Antti H€arm€a, MD, Outi Aikio, MD, PhD, Marita Valkama, MD, PhD,
Markku Leskinen, MD, PhD, Timo Saarela, MD, PhD, and Mikko Hallman, MD, PhD

Objective To study the biologic effect of paracetamol, an inhibitor of prostaglandin synthase, on early closure of
ductus arteriosus, and to evaluate possible adverse effects associated with the drug.
Study design In a controlled, double-blind, phase I-II trial, very low gestational age (<32 weeks) infants requiring
intensive care were randomly assigned to intravenous paracetamol or placebo (0.45% NaCl). A loading dose of
20 mg/kg was given within 24 hours of birth, followed by 7.5 mg/kg every 6 hours for 4 days. Daily cardiac ultrasound
examinations of ductal calibers were performed before the first dose, and until 1 day after the last dose. The main
outcome was a decrease in the ductal caliber without side effects.
Results Of 63 screened infants, 48 were randomized: 23 were assigned to paracetamol and 25 to placebo. Before
the intervention, their ductal calibers were similar. During the intervention, the ductus closed faster in the paracet-
amol group (hazard ratio 0.49, 95% CI 0.25-0.97, P = .016). The mean (95% CI) postnatal ages for ductal closure
were 177 hours (31.1-324) for the paracetamol-treated vs 338 hours (118-557) for controls (P = .045). Paracetamol
serum levels were within the therapeutic range, and no adverse effects were evident.
Conclusions Prophylactic paracetamol induced early closure of the ductus arteriosus without detectable side
effects. Further trials are required to determine whether intravenous paracetamol may safely prevent symptomatic
patent ductus arteriosus. (J Pediatr 2016;-:---).
Trial registration ClinicalTrials.gov: NCT01938261; European Clinical Trials Database: EudraCT 2013-008142-33.

See editorial, p 

P
atent ductus arteriosus (PDA) remains a major clinical challenge in the treatment of very low gestational age infants
(VLGA; born at <32 gestational weeks).1,2 Although spontaneous ductal closure is evident in many VLGA infants,
in others, a hemodynamically significant PDA is associated with severe morbidity.2,3 The optimal time of ductal
closure and the need for medical closure of the ductus remains controversial.4,5 Current therapeutic options include
ibuprofen or indomethacin, which prevent prostaglandin synthesis by inhibiting the cyclooxygenase (COX) enzyme. Sur-
gical closure is used when COX inhibitors are contraindicated or ineffective.6 However, all available therapies have serious
adverse effects.7,8
One strategy to prevent a hemodynamically significant PDA is early therapeutic closure. Previous trials of prophylactic indo-
methacin and ibuprofen medications decreased the risk of PDA, pulmonary hemorrhage, and severe intraventricular hemor-
rhage. However, overall morbidity and disability failed to decrease,9-12 and the risk of pulmonary hypertension and
gastrointestinal bleeding increased.
Paracetamol (acetaminophen) inhibits the peroxidase moiety of the prostaglandin synthase enzyme, decreasing prosta-
glandin synthesis.13 Intravenous paracetamol was introduced in an attempt to reduce the requirement for opiates during res-
piratory treatment after birth.14 The efficacy and safety in the treatment of PDA have recently been studied.15-17 In an
observational study including 201 VLGA infants, paracetamol therapy was associated with a decrease in the incidence of he-
modynamically significant PDA.18 However, the actual biologic drug effect on
early ductal closure after very preterm birth has not been demonstrated.
We hypothesized that early intravenous paracetamol accelerates the From the PEDEGO Research Center, and MRC Oulu,
University of Oulu, and the Department of Children and
contraction of the ductus without serious side effects significantly. To that Adolescents, Oulu University Hospital, Oulu, Finland
end, we designed a randomized, double-blind trial to study the biologic effect Supported by the Alma and K.A. Snellman Foundation,
Oulu, Finland, The Finnish Medical Foundation, The
Foundation for Paediatric Research, and Sigrid Juselius
Foundation. The authors declare no conflicts of interest.
Portions of this study were presented as an abstract at
COX Cyclooxygenase the 30th International Workshop on Surfactant Replace-
ment, Stockholm, Sweden, June 5-6, 2015; 1st Congress
HR Hazard ratio of joint European Neonatal Societies (jENS), Budapest,
LA/Ao Left atrium to aorta ratio Hungary, September 16-20, 2015; and Pediatric Aca-
NICU Neonatal intensive care unit demic Societies (PAS), Baltimore, MD, April 30-May 3,
2016.
PDA Patent ductus arteriosus
VLGA Very low gestational age 0022-3476/$ - see front matter. ª 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jpeds.2016.04.066

1
THE JOURNAL OF PEDIATRICS  www.jpeds.com
of intravenous paracetamol on closure of the ductus arterio-
sus during the first neonatal days in VLGA infants.
Methods
We conducted a randomized, double-blinded trial (The Pre-
mature Infants’ Paracetamol Study, PreParaS) in the Oulu
University Hospital neonatal intensive care unit (NICU)
(ClinicalTrials.gov: NCT01938261; European Clinical Trials
Database: EudraCT 2013-008142-33). The present trial is a
phase I-II study with the aim of establishing a new paraceta-
mol indication in high-risk preterm infants. The study was
performed in accordance with Good Clinical Practice guide-
lines.19 The hospital ethics committee and Finnish Medicines
Agency (Fimea) approved the protocol. The monitoring offi-
cer oversaw the trial arrangements regularly.
All VLGA infants admitted to the NICU were screened.
The duration of gestation was defined by ultrasound exami-
nation before an estimated 16 weeks of pregnancy. Exclusion
criteria were septic shock, major malformation, and chromo-
somal abnormality. The study doctors obtained written
informed consent from parents within 24 hours of birth.
Infants were assigned randomly to intravenous paraceta-
mol (ie, Perfalgan 10 mg/mL [Oy Bristol-Myers Squibb
Finland Ab, Espoo, Finland], or Paracetamol Fresenius
Kabi 10 mg/mL [Fresenius Kabi Ab, Helsinki, Finland] solu-
tions for infusion) or placebo (0.45% saline solution) groups.
The paracetamol preparation was changed during the study
period because of hospital drug policy. The investigators
had no influence on the drug choice.
Computed randomization was performed using a 4-block
design. To decrease the risk of significant heterogeneity be-
tween cases and controls, individual treatment strata were
defined by sex and gestational age. The treatment allocation
codes were sealed in sequentially labeled opaque envelopes.
Both paracetamol and saline solutions appeared equally
transparent in the syringe. All nurses and doctors involved
in the treatment and study of the infants were blinded to
the study medication.
A separate team of nurses prepared the study drug in a
study pharmacy outside NICU. The drug was given to the
study patient’s nurse in a syringe. The dose of paracetamol
was equal to that used for pain therapy in neonates,14 the
loading dose being 20 mg/kg, followed by the maintenance
dose of 7.5 mg/kg every 6 hours for 4 days, given as 15-min-
ute infusions. No other paracetamol preparations were given
before, during, or 2 days after (washout period) the study
medication.
The first cardiac ultrasound examination was performed
before the study drug, and then once a day until 1 day after
the study medication period. Thereafter, infants with an
open ductus were examined 1-2 times per week, unless other-
wise indicated. All participants were studied for patency of
the ductus at discharge from NICU.
Cardiac ultrasound examinations were performed by neo-
natologists who underwent training by the pediatric cardiol-
ogist to enhance the uniformity of evaluations. Patients were
2 €rkin et al
Volume -
examined in the supine position with slight left shoulder
recumbences. The recordings were obtained using Vivid
i (first 8 patients) or Vivid S5 (GE Healthcare, Helsinki,
Finland) ultrasound devices with appropriate transducers
(12S or 6S). No sedation was used during the ultrasound ex-
aminations. In cases of distress, the neonatal nurse soothed
the infant, and oral 20% glucose solution drops were admin-
istered via a pacifier if necessary.
Daily ultrasound measurements included ductal caliber
(mm) from the narrowest point of the vessel, and left atrium
to aorta ratio (LA/Ao) from M-mode. The smallest internal
diameter of the ductal pulmonary end was measured from
the parasagittal plane of the high left parasternal window,
with the transducer notch at noon (ductal view). Ductal
caliber was calculated in millimeters per kilogram of birth
weight (mm/kg) as well. Left atrial systolic dimensions were
measured at the greatest distance between the anterior aspect
of the left atrial posterior wall and the inner aortic posterior
wall. The aorta was measured from the anterior surface of the
anterior root echo to the anterior surface to the posterior root
echo.
Paracetamol concentrations were analyzed from serum
samples stored at 70 C using the Paracetamol Assay Kit
(Cambridge Life Sciences Ltd, Ely, United Kingdom). All sam-
ples were analyzed as duplicates. Occasional sample turbidity
meant that sample blanks without color reagent served as the
reference. The interassay and intra-assay coefficients of varia-
tion were 18.1% and 14.1%, respectively. No change in para-
cetamol concentration was detected after prolonged storage.
Primary and Secondary Outcomes
The primary outcome of the trial was the decrease in and
closure of the ductus during the intervention as function of
postnatal age. The secondary outcomes were LA/Ao, the
age of permanent closure of the ductus, ductus therapies,
the side effects of paracetamol, and neonatal and long-term
morbidity and mortality.
Adverse events were assessed by continuously monitoring
the infants’ symptoms and signs (eg, oxygenation, blood
pressures, and inotrope use) and laboratory values (eg, plate-
lets, serum sodium, and bilirubin). Renal function was moni-
tored by measurements of diuresis (mL/kg/h). Phototherapy
was administered by serum bilirubin level using a specific
nomogram based on the duration of gestation and postnatal
age. The symptoms of pain and discomfort were assessed us-
ing pain scales, and the need for intravenous morphine was
judged clinically.20
The diagnosis of hemodynamically significant PDA
included the clinical criteria of cardiopulmonary distress
(increased need for respiratory support, decreased systolic
or mean blood pressure, increased pulse pressure, pulmonary
congestion, cardiomegaly, hepatomegaly, a murmur, hyper-
dynamic precordium, or bounding pulses) and the following
echocardiography criteria: LA/Ao >1.4, PDA diameter >50%
wider than left pulmonary artery, and the flow patterns
showing a large volume left-to-right ductal shunt.5,21 Treat-
ment options included intravenous ibuprofen and surgical
Ha
- 2016 ORIGINAL ARTICLES

ligation. Bronchopulmonary dysplasia,22 intraventricular

hemorrhage,23 necrotizing enterocolitis,24 and retinopathy Table I. Baseline clinical characteristics
of prematurity25 were defined as described previously. Paracetamol Placebo group
group (n = 23) (n = 25)
Statistical Analyses Gestational age, wk, mean (SD) 28.4 (2.36) 28.3 (2.06)
Based on the previous studies, it was assumed that the ductus Gestational age <29 wk, n (%) 11 (48) 11 (44)
Gestational age <27 wk, n (%) 4 (17) 4 (16)
caliber would decrease in 50% of the paracetamol-treated in- Birth weight, kg, mean (SD) 1.22 (0.43) 1.12 (0.34)
fants vs 10% of untreated infants.16,18 Therefore, the sample Birth weight Z-score, mean (SD) 0.97 (1.34) 1.33 (1.36)
size, resulting in the 40% proportional difference between Antenatal steroids, n (%) 20 (87) 25 (100)
Antenatal steroid doses/mother, mean (SD) 1.65 (0.89) 1.64 (0.49)
paracetamol-treated and controls, was calculated as 48 (po- PPROM, n (%) 10 (43) 6 (24)
wer 80%, alpha error 0.05). PPROM, d, median (range) 0 (0-34) 0 (0-35)
All outcomes were analyzed by intention to treat. For Cesarean delivery, n (%) 15 (65) 10 (40)
Male sex, n (%) 13 (57) 14 (56)
continuous variables, the independent samples t-test or the Twins, n (%) 7 (30) 10 (40)
Mann-Whitney U-test, and for categorical values, the c2 Apgar scores at 1 min, median (range) 7 (2-9) 6 (1-9)
test, were used as appropriate. The ductal closures as a func- Apgar scores at 5 min, median (range) 7 (4-10) 7 (2-9)
Surfactant (100 mg/kg), n (%) 16 (70) 22 (88)
tion of postnatal age were estimated using Kaplan-Meier Surfactant doses/patient, mean (SD) 1.3 (1.3) 1.4 (0.9)
analysis, and the hazard ratios (HRs) using the Cox regres-
PPROM, prolonged premature rupture of the fetal membranes.
sion analysis. A repeated measures ANOVA was used to
assess the daily variation in ductal and fluid balance measure- According to post hoc analysis, 13 infants (27.1%) had an
ments between the groups. An interrater reliability analysis open ductus 1 day after the study medication period (post-
using the Cronbach alpha statistic was performed to deter- natal ages 120-144 hours): 4 in the paracetamol group and
mine consistency among raters of cardiac ultrasound. The nine in the placebo group. No PDA treatment took place dur-
statistical analyses were performed using IBM SPSS Statistics ing study drug administration. Seven infants (14.6%)
for Windows, Version 22.0 (IBM Corp, Armonk, New York). required therapy for PDA before discharge from the NICU
P < .05 was considered significant. (Table III). As expected, gestational age influenced the
ductal closure. For infants born after 27 weeks of gestation,
Results the mean (SD) postnatal age for ductal closure was
80 hours (151) in the paracetamol group vs 322 hours
A total of 63 VLGA infants were screened for the study be- (514) in the placebo group (P = .004). For the extremely
tween September 18, 2013, and January 2, 2015 (Figure 1; preterm infants (born at <27 weeks gestation, n = 8), an
available at www.jpeds.com). Of these, 48 underwent acute paracetamol effect on the contraction of ductus
randomization: 23 infants assigned to the paracetamol arteriosus was not detected (P = .63), and 4 (50%) required
group and 25 to the placebo group. Baseline clinical PDA treatment (paracetamol n = 3, placebo n = 1).
characteristics were similar in both groups (Table I). Paracetamol apparently increased closure of ductus in boys
The protocol of serial cardiac ultrasound examinations (HR 0.31, 95% CI 0.12-0.85, P = .023) and not in girls (HR
(n = 330) was performed for all the study infants. On the ba- 0.72, 95% CI 0.27-1.96, P = .52).
sis of 3 measurements (ductal caliber, LA/Ao, left pulmonary Table III shows adverse events and neonatal outcomes.
artery caliber) for the 3 raters, we calculated the interrater During the study medication period, no differences in
reliability: alpha = 0.97 (95% CI 0.90-0.99). Before the onset adverse events were detected between the infants treated
of the study medication, the mean (SD) ductal caliber was with paracetamol or placebo. During the first week of life,
1.57 (0.66) mm in the paracetamol group vs 1.39 (0.76) the groups had similar diuresis rates (P = .102; Table II)
mm in the placebo-group (P = .38). The daily ductal sizes and frequency of hypernatremic serum sodium values
(mm, mm/kg) and LA/Aos are listed in Table II (available (>150 mmol/L; Table III). There was no evidence of
at www.jpeds.com). The variation was considerable, and no paracetamol-induced hypotension because the requirement
differences between the 2 groups were detectable. of inotropes was similar (P = .74; Table III). Two patients
The frequencies of a closed ductus arteriosus as a func- treated with paracetamol (23+5 and 28+5 gestation weeks)
tion of postnatal age are shown in Figure 2. In the with histories of prolonged premature rupture of fetal
paracetamol group, the closure rate tended to be faster membranes had severe respiratory distress with pulmonary
and at end of the intervention, the number of infants hypertension. The first child developed pulmonary
with a closed ductus arteriosus tended to be higher (HR hypertension before paracetamol and the other after onset
0.49, 95% CI 0.25-0.97, P = .016). The median postnatal of paracetamol. Both presented acute favorable responses
ages for the observed ductal closure were 41 hours (IQR, to inhaled nitric oxide. No difference in the highest
33-85 hours) for the paracetamol-treated group vs fractions of supplemental oxygen during the study
78 hours (IQR, 50-375 hours) for the controls, and the medication was evident (P = .85; Table III), and no signs
means (SD) were 177 hours (338) vs 336 (517), of hepatotoxicity were observed. One infant who was
respectively (P = .045). There were no detectable treated with placebo died of necrotizing enterocolitis at the
differences in the LA/Aos (P = .31; Table II). age of 28 days.
Paracetamol Accelerates Closure of the Ductus Arteriosus after Premature Birth: A Randomized Trial 3
THE JOURNAL OF PEDIATRICS  www.jpeds.com
Figure 2. Kaplan-Meier survival curve for ductal closure. Percentage of paracetamol and control group infants with open ductus
up to 144 hours after birth.
Paracetamol concentrations were analyzed in 87 serum
samples obtained from 21 infants during paracetamol
dosage, the observed concentrations ranging from undetect-
able to 25.2 mg/L. No significant accumulation of paraceta-
mol was evident and the concentrations decreased as a
function of time. Four to 6 hours after paracetamol infusions,
mean (SD) concentrations were 4.6 (4.5) mg/L. The mean
(SD) concentrations during day one, 9.8 (5.5) mg/L
(n = 26), were not different from those during day four,
11.8 (6.0) mg/L (n = 17, P = .29). No sex difference was
evident: boys 9.2 (6.2) mg/L vs girls 9.0 (5.6) mg/L
(P = .91). No difference between infants born before or after
29 gestation weeks was detected: mean (SD) concentration
9.4 (5.5) vs 8.9 (6.2) mg/L, respectively (P = .71). Therefore,
paracetamol concentrations did not explain the differences in
the ductal caliber results.
Discussion
In this randomized trial, intravenous paracetamol potenti-
ated the early closure of the ductus arteriosus after very
preterm birth. In most infants, the ductus arteriosus con-
tracted within 3 days during paracetamol treatment. Un-
like other early treatments of the ductus arteriosus using
surgery or COX inhibitors, paracetamol seemed to be
well-tolerated.7,8
4 €rkin et al
Volume -
A hemodynamically significant shunt from systemic to
pulmonary circulation through a PDA contributes to symp-
toms and adverse outcomes in infants born very preterm.2
There is large individual variation in the closure after birth
and in the magnitude of the shunt through a PDA. Because
immature infants are predisposed to cardiac decompensa-
tion, early medical closure of the ductus arteriosus is an
attractive approach. Unfortunately, the early closure of the
ductus using surgery or COX inhibitors has been associated
with serious adverse effects.9-12,26 The present paracetamol
trial was based on the previous use of intravenous paraceta-
mol to limit the use of opiates and their adverse effects during
respiratory therapy after very preterm birth.14 Unexpectedly,
our retrospective cohort study revealed an association be-
tween early intravenous paracetamol and a decrease in the
incidence of PDA.18 Recent studies have also shown that
paracetamol treatment after the first postnatal week induced
the closure of PDA without apparent adverse effects.15,16,27,28
In our study, the effect of intravenous paracetamol on the
contraction of the ductus was limited to the male sex and to
those born after 27 weeks of gestation. We were unable to
detect any sex- or gestational age-related differences in serum
paracetamol levels. Because the number of subjects was small
and the duration of treatment remained limited to 4 days, the
population effect may prove to be more general than
observed in the present study.
Ha
- 2016 ORIGINAL ARTICLES

potential effect on PDA. However, in a case series,

Table III. Secondary outcomes paracetamol-induced early ductal closures were described
Paracetamol Placebo after much higher doses.32
group group
(n = 23) (n = 25)
Data on the safety of paracetamol for VLGA infants has
been reported.31,33 Similar to previous reports, we found lit-
Persistent ductus arteriosus, treated, n (%) 4 (17) 3 (12)
Pulmonary outcomes tle evidence of side effects, including hepatic toxicity or hypo-
Mechanical ventilation, d, median (range) 1 (0-32) 2 (0-39) tension (Table III). Although there is some evidence about
Supplemental oxygen, d, mean (SD) 20.0 (24.5) 22.4 (25.0) paracetamol-induced decrease on blood pressure, we found
Maximal fraction of supplemental 0.45 (0.45) 0.46 (0.23)
oxygen, mean (SD) no difference in the inotrope requirements between the 2
Adverse events groups.34,35 The highest level of observed paracetamol
Inotrope requirement, n = 0/1/2/3* 17/7/0/1 16/5/0/3 concentration, 25.2 mg/L, was considerably lower than that
Oliguria (<1 mL/kg/h), n (%) 5 (22) 7 (28)
Polyuria (>5 mL/kg/h), n (%) 6 (26) 9 (36) reported in accidental paracetamol poisoning (117-180 mg/
Hypernatremia (>150 mmol/L), n (%) 5 (22) 12 (48) L).36 However, more study is needed to confirm the safety
Phototherapy periods, n, mean (SD) 3.5 (1.7) 3.9 (2.0) of paracetamol in this vulnerable population.
Highest serum bilirubin, mmol/L, mean (SD) 159 (21.2) 158 (18.7)
Clinical infection, n (%) 9 (39) 10 (40) In addition to their effect on the ductus, COX inhibitors
Septicemia: positive blood culture, n (%) 4 (17) 3 (12) cause more generalized vasoconstriction than paracetamol,
Neonatal outcomes and may cause serious side effects, including renal insuffi-
BPD grade 1: supplemental 7 (30) 11 (44)
oxygen at 28 days, n (%) ciency, gastrointestinal perforation, intraventricular hemor-
BPD grades 2-3: supplemental oxygen 0 1 (4) rhage, and pulmonary hypertension.9 As a peroxidase
at 36 wk after conception, n (%) inhibitor, paracetamol seems to have fewer side effects than
Retinopathy, resolved, n (%) 2 (9) 2 (8)
Retinopathy, treated, n (%) 1 (4) 0 COX inhibitors. However, in addition to inhibition of pros-
Intraventricular hemorrhage, 5 (22) 9 (36) taglandin synthesis, paracetamol may also have other bio-
grades 1-2, n (%) logic effects that potentially influence the immature infant.
Intraventricular hemorrhage, 1 (4) 1 (4)
grades 3-4, n (%) In the present study, 1 infant with prolonged premature
Necrotizing enterocolitis, stage 3, n (%) 0 1 (4) rupture of fetal membranes developed pulmonary hyperten-
Death, n (%) 0 1 (4) sion after the onset of paracetamol. Although prolonged pre-
BPD, bronchopulmonary dysplasia. mature rupture of fetal membranes and very preterm birth
*Calculated inotropes: 0 = none given, 1 = dopamine or dobutamine, 2 = dopamine and dobut- are associated with acute pulmonary hypertension,37 the pos-
amine, 3 = dopamine, dobutamine, and norepinephrine.
sibility that paracetamol may promote pulmonary hyperten-
The risk of PDA is associated with lack of antenatal ste- sion remains to be investigated further.
roids, hereditary predisposition, and particularly with the de- The present study has some limitations. The size of this
gree of immaturity.1 In the fetus, the ductus arteriosus is kept study was calculated to be sufficient to reveal an acute bio-
actively open, and it closes after term birth in 2 phases: func- logic effect of paracetamol on the ductus arteriosus. The re-
tional closure within the first hours after birth, followed by sults show a robust biologic effect that may define a new
anatomic occlusion over the next several days.29 Acute func- potential indication for an established medication. However,
tional closure is achieved by contraction of ductal wall this study had insufficient power to prove an effect in sub-
smooth muscle cells sensitive to vasodilator prostaglandins. groups, establish drug safety, or confirm efficacy in
Functional closure is reinforced by platelets that form clots decreasing the incidence of hemodynamically significant
within the contracting vessel.30 In contrast with COX inhib- PDA. It included mostly larger infants for whom treatment
itors, paracetamol is not reported to inhibit platelet aggrega- strategies for PDA are less uncertain. Finally, the dose re-
tion, which may prove to be a benefit of paracetamol for quirements for infants born extremely preterm need to be
ductal closure in high-risk infants. investigated further.
The study medication was given intravenously to pre- In conclusion, prophylactic intravenous paracetamol,
term infants known to have a low rate of paracetamol elim- compared with placebo, accelerated closure of the ductus ar-
ination.31 The dosage used was according to the teriosus in VLGA infants without detectable adverse effects,
recommendations of pharmacokinetic studies (first day providing evidence for its biologic effect and safety. Because
42.5 mg/kg, then 30 mg/kg/d) and lower than that used paracetamol may serve additionally as a nonsedative anal-
for PDA closure after the first week of life (60 mg/kg/ gesic, it could be an attractive alternative drug for opiates
d).15,31 We measured paracetamol serum levels from 21 in- and COX inhibitors. Further studies, including large ran-
fants and obtained results that were consistent with previ- domized trials, are required to define the clinical potential
ous data.31 There was no evidence on either the and limitations of paracetamol for premature infants. n
accumulation or excessive elimination of paracetamol.
We found no effect on the risk of PDA. This was antici- Submitted for publication Jan 13, 2016; last revision received Mar 16, 2016;
accepted Apr 20, 2016.
pated because the trial was not planned to detect a change
Reprint requests: Outi Aikio, MD, PhD, Department of Children and
in clinical outcome, and the contemporary use of intrave- Adolescents, Oulu University Hospital, PO Box 23, 90029 OYS, Finland.
nous paracetamol after the trial may also dilute any E-mail: outi.aikio@ppshp.fi

Paracetamol Accelerates Closure of the Ductus Arteriosus after Premature Birth: A Randomized Trial 5
THE JOURNAL OF PEDIATRICS  www.jpeds.com Volume -

20. Polkki T, Korhonen A, Axelin A, Saarela T, Laukkala H. Development

References and preliminary validation of the Neonatal Infant Acute Pain Assess-
ment Scale (NIAPAS). Int J Nurs Stud 2014;51:1585-94.
1. Hamrick SE, Hansmann G. Patent ductus arteriosus of the preterm in-
21. Skinner J, Alverson D, Hunter S. Echocardiography for the neonatolo-
fant. Pediatrics 2010;125:1020-30.
gist. London: Churchill Livingstone; 2000.
2. Noori S, McCoy M, Friedlich P, Bright B, Gottipati V, Seri I, et al. Failure
22. Ehrenkranz RA, Walsh MC, Vohr BR, Jobe AH, Wright LL,
of ductus arteriosus closure is associated with increased mortality in pre-
Fanaroff AA, et al. Validation of the National Institutes of Health
term infants. Pediatrics 2009;123:e138-44.
consensus definition of bronchopulmonary dysplasia. Pediatrics 2005;
3. Rolland A, Shankar-Aguilera S, Diomande D, Zupan-Simunek V,
116:1353-60.
Boileau P. Natural evolution of patent ductus arteriosus in the extremely
23. Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolu-
preterm infant. Arch Dis Child Fetal Neonatal Ed 2015;100:F55-8.
tion of subependymal and intraventricular hemorrhage: a study of
4. Clyman RI, Chorne N. Patent ductus arteriosus: evidence for and against
infants with birth weights less than 1,500 gm. J Pediatr 1978;92:
treatment. J Pediatr 2007;150:216-9.
529-34.
5. Zonnenberg I, de WK. The definition of a haemodynamic significant
24. Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al.
duct in randomized controlled trials: a systematic literature review.
Neonatal necrotizing enterocolitis. Therapeutic decisions based upon
Acta Paediatr 2012;101:247-51.
clinical staging. Ann Surg 1978;187:1-7.
6. Heuchan AM, Clyman RI. Managing the patent ductus arteriosus: cur-
25. International Committee for the Classification of Retinopathy of Prema-
rent treatment options. Arch Dis Child Fetal Neonatal Ed 2014;99:
turity. The International Classification of Retinopathy of Prematurity re-
F431-6.
visited. Arch Ophthalmol 2005;123:991-9.
7. Mirea L, Sankaran K, Seshia M, Ohlsson A, Allen AC, Aziz K, et al. Treat-
26. Fowlie PW, Davis PG, McGuire W. Prophylactic intravenous indometh-
ment of patent ductus arteriosus and neonatal mortality/morbidities:
acin for preventing mortality and morbidity in preterm infants. Co-
adjustment for treatment selection bias. J Pediatr 2012;161:689-94.
chrane Database Syst Rev 2010;(7):CD000174.
8. Sosenko IR, Fajardo MF, Claure N, Bancalari E. Timing of patent ductus
27. Ohlsson A, Shah PS. Paracetamol (acetaminophen) for patent ductus ar-
arteriosus treatment and respiratory outcome in premature infants: a
teriosus in preterm or low-birth-weight infants. Cochrane Database Syst
double-blind randomized controlled trial. J Pediatr 2012;160:929-35.
Rev 2015;3:CD010061.
9. Gournay V, Roze JC, Kuster A, Daoud P, Cambonie G, Hascoet JM, et al.
28. Oncel MY, Yurttutan S, Degirmencioglu H, Uras N, Altug N, Erdeve O,
Prophylactic ibuprofen versus placebo in very premature infants: a rand-
et al. Intravenous paracetamol treatment in the management of patent
omised, double-blind, placebo-controlled trial. Lancet 2004;364:1939-44.
ductus arteriosus in extremely low birth weight infants. Neonatology
10. Kluckow M, Jeffery M, Gill A, Evans N. A randomised placebo-
2012;103:165-8.
controlled trial of early treatment of the patent ductus arteriosus. Arch
29. Clyman RI. Mechanisms regulating the ductus arteriosus. Biol Neonate
Dis Child Fetal Neonatal Ed 2014;99:F99-104.
2006;89:330-5.
11. Ohlsson A, Shah SS. Ibuprofen for the prevention of patent ductus arte-
30. Echtler K, Stark K, Lorenz M, Kerstan S, Walch A, Jennen L, et al. Plate-
riosus in preterm and/or low birth weight infants. Cochrane Database
lets contribute to postnatal occlusion of the ductus arteriosus. Nat Med
Syst Rev 2011;(7):CD004213.
2010;16:75-82.
12. Schmidt B, Davis P, Moddemann D, Ohlsson A, Roberts RS, Saigal S,
31. Allegaert K, Palmer GM, Anderson BJ. The pharmacokinetics of intrave-
et al. Long-term effects of indomethacin prophylaxis in extremely-
nous paracetamol in neonates: size matters most. Arch Dis Child 2011;
low-birth-weight infants. N Engl J Med 2001;344:1966-72.
96:575-80.
13. Anderson BJ. Paracetamol (acetaminophen): mechanisms of action.
32. El-Khuffash A, Jain A, Corcoran D, Shah PS, Hooper CW, Brown N,
Paediatr Anaesth 2008;18:915-21.
et al. Efficacy of paracetamol on patent ductus arteriosus closure may
14. Harma A, Aikio O, Hallman M, Saarela T. Intravenous paracetamol de-
be dose dependent: evidence from human and murine studies. Pediatr
creases requirements of morphine in very preterm infants. J Pediatr
Res 2014;76:238-44.
2016;168:36-40.
33. Allegaert K, Rayyan M, De RT, Van BF, Naulaers G. Hepatic tolerance of
15. El-Khuffash A, James AT, Cleary A, Semberova J, Franklin O, Miletin J.
repeated intravenous paracetamol administration in neonates. Paediatr
Late medical therapy of patent ductus arteriosus using intravenous para-
Anaesth 2008;18:388-92.
cetamol. Arch Dis Child Fetal Neonatal Ed 2015;100:F253-6.
34. Allegaert K, Naulaers G. Haemodynamics of intravenous paracetamol in
16. Hammerman C, Bin-Nun A, Markovitch E, Schimmel MS, Kaplan M,
neonates. Eur J Clin Pharmacol 2010;66:855-8.
Fink D. Ductal closure with paracetamol: a surprising new approach
35. Chiam E, Weinberg L, Bailey M, McNicol L, Bellomo R. The haemody-
to patent ductus arteriosus treatment. Pediatrics 2011;128:e1618-21.
namic effects of intravenous paracetamol (acetaminophen) in healthy
17. Oncel MY, Yurttutan S, Erdeve O, Uras N, Altug N, Oguz SS, et al. Oral
volunteers: a double-blind, randomized, triple crossover trial. Br J Clin
paracetamol versus oral ibuprofen in the management of patent ductus
Pharmacol 2016;81:605-12.
arteriosus in preterm infants: a randomized controlled trial. J Pediatr
36. Porta R, Sanchez L, Nicolas M, Garcia C, Martinez M. Lack of toxicity
2014;164:510-4.
after paracetamol overdose in an extremely preterm neonate. Eur J
18. Aikio O, Harkin P, Saarela T, Hallman M. Early paracetamol treatment
Clin Pharmacol 2012;68:901-2.
associated with lowered risk of persistent ductus arteriosus in very pre-
37. Aikio O, Metsola J, Vuolteenaho R, Perhomaa M, Hallman M. Transient
term infants. J Matern Fetal Neonatal Med 2014;27:1252-6.
defect in nitric oxide generation after rupture of fetal membranes and
19. European Medicines Agency. Guideline for Good Clinical Practice.
responsiveness to inhaled nitric oxide in very preterm infants with hyp-
http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_
oxic respiratory failure. J Pediatr 2012;161:397-403.
guideline/2015/08/WC500191488.pdf. July 23, 2015.

6 €rkin et al
Ha
- 2016 ORIGINAL ARTICLES

63 Patients assessed for eligibility

15 Excluded
4 Did not meet eligibility criteria
7 Consent denied by parents
4 Other reasons (consent was not
sought, language difficulties)

48 Randomized

23 Randomized to paracetamol 25 Randomized to placebo

23 Included in the primary analysis 25 Included in the primary analysis

Figure 1. CONSORT diagram.

Paracetamol Accelerates Closure of the Ductus Arteriosus after Premature Birth: A Randomized Trial 6.e1
6.e2

THE JOURNAL OF PEDIATRICS

Table II. Daily cardiac ultrasound measurements, diuresis, and serum sodium levels during days 0-5 after birth
Day 0* Day 1 Day 2 Day 3 Day 4 Day 5*
0-24 h 25-48 h 49-72 h 73-96 h 97-120 h 121-144 h
Paracet Placebo Paracet Placebo Paracet Placebo Paracet Placebo Paracet Placebo Paracet Placebo P value†
Ductal caliber, mm 1.57 (0.66) 1.39 (0.76) 0.73 (0.90) 1.00 (0.85) 0.48 (0.81) 0.80 (0.74) 0.24 (0.48) 0.58 (0.74) 0.17 (0.40) 0.35 (0.49) 0.21 (0.48) 0.38 (0.54) .084
Ductal caliber, mm/kg 1.35 (0.61) 1.32 (0.73) 0.74 (0.92) 0.88 (0.66) 0.53 (0.94) 0.76 (0.74) 0.34 (0.67) 0.55 (0.69) 0.25 (0.58) 0.33 (0.50) 0.28 (0.63) 0.40 (0.60) .356
LA/Ao 1.20 (0.18) 1.24 (0.29) 1.22 (0.23) 1.34 (0.39) 1.19 (0.25) 1.20 (0.31) 1.20 (0.26) 1.18 (0.33) 1.20 (0.24) 1.22 (0.32) 1.17 (0.36) 1.23 (0.26) .310
Diuresis, mL/kg/h 1.76 (0.98) 2.02 (1.10) 3.48 (1.60) 4.18 (1.17) 3.10 (1.00) 3.86 (1.05) 3.19 (0.96) 3.32 (1.08) 2.87 (1.23) 2.80 (1.27) 3.46 (1.52) 3.02 (1.30) .102
Serum sodium, mmol/L 139 (2.33) 140 (2.73) 143 (3.47) 143 (3.23) 146 (3.56) 146 (2.51) 146 (4.55) 147 (4.72) 144 (3.73) 146 (5.41) 144 (3.13) 144 (5.37) .604

Paracet, paracetamol group; Placebo, placebo group.


Values are mean (SD).

www.jpeds.com
*Day 0 measurements represent the data before the study medication, and day 5 measurements were obtained one day after the study medication.
†The difference between the 2 groups’ daily measurements from days 1-5, repeated measures ANOVA was used.

Volume -
Ha
€rkin et al